http://www.youtube.com/watch?v=0_TC20cBZLA http://www.youtube.com/watch?v=_Yas_Cl2UvM https://arstechnica.com/gaming/2018/10/segas-genesis-and-more-get-an-hdmi-upgrade-with-the-mega-sg/ It's happening! We've got a sub-$200 FPGA Genesis, with expansion port for Sega CD, and $10 cartridge adapters for Master System/Mark III, SG-1000, and Game Gear. 32X support is still being worked on, but this is looking to be the nigh-all-encompassing solution for pre-Saturn Sega goodness in the HD age.
Very cool. I like the design of the system itself, especially the JP color variant. The most exciting part to me is honestly the 8BitDo M30 controller. I like that they changed the contours on the top and bottom and added shoulder buttons. Kinda looks like a hybrid of the Genesis and Saturn Model 2/JP controller. Very curious to see how it compares to Saturn pads. My guess is it will feel better than original Genesis/Mega Drive controllers (both 3 button and 6 button ones), but not quite as good as original Saturn Model 2/JP controllers. (That said, I've seen quite a few complaints about the d-pad for their SN30 controller. I use original SNES pads with my Super Nt so I haven't tried the SN30 out myself. Hopefully there aren't any similar quality issues with the M30. Definitely picking one up since the Genesis/MD controller--both the 3 button and 6 button--is hard to go back to, especially compared to the Saturn pad which is the best game controller ever made for 2D games. But I still have some 6 button Genesis pads in good condition so if the M30 ends up being a bust I'll at least have a decent, authentic option.) I do also wonder how far off we are from an FPGA retro system that is able to output at 4K resolution at the same $189 price point. Kevtris has said it isn't commercially viable yet, and that certainly seems to be the case. But I would assume costs will come down over time, and that it could be viable in 5-10 years. That's a long enough time to have to wait that I'm still definitely getting a Mega Sg now, but once they come out with new models of the Super Nt and Mega Sg that have new hardware that enables 4K video output, that'll be the be-all-end-all for playing 8 and 16 bit games in HD resolution since the resolution those games run at will scale better to 4K than they do to 1080p. Looking forward to when that day comes!
Doesn't look so accurate when it's missing some borders on the sides. EDIT: For reference: http://www.youtube.com/watch?v=OZxIz3ivjZY Also, it seems to be missing those moving dots in the lower border?
WANT! WANT! WANT! WANT! WANT! Since I've never owned a Sega system in my life, this is my perfect opportunity to finally get one. Initially I was going to get a Sega Genesis as a child in 1998, but I was told at the Best Buy that the console was discontinued. ;_; Sega's support for PC in the mid to late '90s made up for it though, which caused me to lose interest in the system.
It's been pretty fun seeing journalists unfamiliar with Sega trying to report on this thing. I've seen multiple outlets refer to the original system simply as a "Sega," and Kotaku claimed that it would play, "Genesis, Mega Drive, and Master Drive" games.
I wish they'd show a picture with the Sega CD connected. I want to see how weird it looks given the size difference.
Pre-ordered this as soon as I heard about it and their site came back up. Also pre-ordered 2 of the wireless 8Bitdo M30 controllers - I can't wait! //forums.sonicretro.org/public/style_emoticons/default/v.png
Why would you want overscan borders when you will have multiple resolution options on the system itself?
Well, the reference footage uses the slightly horizontally-squished NTSC aspect ratio. I imagine that can be toggled on or off, but I suppose for pure authenticity, Having the full border and DMA garbage adds to it. Website's still down, but here's more reference footage from their YouTube channel: http://www.youtube.com/watch?v=gSeI4WqWAPY http://www.youtube.com/watch?v=B94VXX8amAk
It'll look pretty tiny on top of a Model 1 Sega CD, but I think it'll probably look pretty good proportionally on top of a Model 2 Sega CD, like the Model 2 Genesis does. The Mega Sg is 168 mm wide, 138 mm long, and 47 mm high. The Genesis Model 2 is 220 mm wide, 210 mm long, and 60 mm high. (The Genesis Model 1 is 280 mm wide, 215 mm long, and 57 mm high.) So the Mega Sg is obviously quite a bit smaller than both models of the original Genesis/MD, but it's closer in size to the Model 2, especially when it comes to its width. Therefore I think it'll look better with the Sega CD Model 2 than it will the Model 1, especially if you align the front of the Mega Sg along the front of the Model 2 Sega CD (which will leave a decent amount of space for the back-left side of the complete Mega Sg+Sega CD Model 2 unit, but that's fine since it won't be visible from the front and having that space will actually be good for cord management). Can't wait until Linneman and others get their hands on this and are able to give us their full impressions and comparisons. Can we fast-forward 5-10 years so we can just already have the updated hardware revision that enables 4K output, which will be the ultimate endpoint for enjoying 8 and 16 bit era games in HD? Exciting times ahead.
I hope it's accurate enough. I never took the man making these things as being all that familiar with the Mega Drive community. Anyone know if it can run Overdrive and Overdrive 2? I'm rather curious. I hope it's smooth sailing, regardless. It's quite picky hardware to get running well. Regardless, I'd like to grab one. Just a bit broke right now.
most if not all modern tv's allow overscan adjustments like fit or full, so no real need for that frame any more. I understand why people would miss it hell it was even bigger in pal land, megadrive gaming back then felt like widescreen with the top and bottom bars being full fat.
No he's right, an actual, full-fat proper Mega Drive clone should have its cons and well as its pros. Though I don't know if their methods of direct capture produce a different image to what would be rendered on a CRT. I'm not actually sure what those dots are but a quick skim of YouTube suggests they don't appear in the PAL version? Not that I'm pretending to understand what it is and why it appears. Or if indeed it's even a feature shared across different revisions of the hardware. I could really do with a crash course on this subject so I can add things to the wiki. How all these Sega consoles are meant to appear across multiple televisions and video cables.
Overscan borders actually get graphics rendered into them on Overdrive 2, so there's that. :v: Edit for Black Squirrel's benefit: Those are CRAM dots. They're drawn up because of where the game is writing palette entries during that period of the vblank. I presume PAL software would deliberately perform those CRAM writes later, because otherwise they'd show up inside the 50Hz border. 60Hz seems to have its border cropped out by most TVs anyway.
They do actually appear on PAL machines too, the cause is (as Flygon mentioned) writing to CRAM during active display. Although the VDP triggers V-blank on the 68k, the VDP is still writing backdrop colour to display (the boarder colour on the bottom of the screen). So even though the 68k has interrupted into the V-blank routine, the VDP is still writing colour data as it's actually still during display, so any CRAM writes still cause dots to appear during the boarder. The reason you don't see PAL games display it, is because software deliberately delays the CRAM write for a short while to give the TV time to get into the "real" V-blank (when the VDP is not actually writing the backdrop colour). The same software generally (and intentionally) do not give NTSC this same delay simply because NTSC has less time during V-blank to transfer data, although you could delay until the TV actually reaches real V-blank, it would be a heavy loss of time for NTSC, you can only transfer a maximum of about $1D30 (or less if the H/V resolution is shorter) approx bytes via DMA during V-blank, so time is heavily critical, waiting for real V-blank would be foolish on NTSC hence why they didn't do that. For PAL machines, you have roughly $3914 (or less if the H/V resolution is shorter) approx bytes via DMA, so if you can transfer the data in NTSC within V-blank fine, you'll definitely be able to transfer the same data on PAL systems even with deliberately delaying for real V-blank. This is why PAL machines get the delay treatment, and NTSC machines don't, it is software controlled though. As for why the dots appear at all, I do not actually know the exact science, but from what I can work out, during H-blank the VDP is calculating the scanlines of colours to display on the next active display, in order to do that, it needs to read from CRAM during H-blank, but if you are writing new colours into CRAM during H-blank (real TV/VDP H-blank) while the VDP is reading from CRAM, the mechanism (or address) the VDP is meant to be reading from is probably being directed to the FIFO and reading the last value in there as opposed to the CRAM address (again, I don't have any hard evidence so don't take what I say as definite), it's similar to performing a DMA copy of data (copy, not transfer) where the copy destination is the sprite table, the caché doesn't update properly, and instead the VDP ends up copying data from the FIFO instead of the sprite table address in VRAM. These could be related. I have noticed VSRAM to have a similar issue if you are to write too much data into VSRAM during H-blank, it can cause the -1 V-scroll to screw up (not in the common AND bitwise way that it normally does). EDIT: I'm wondering if it's to do with the fact that CRAM and VSRAM are internally within the VDP, you don't get this issue with VRAM writes which is external.
Sweet - thanks for that Turns out Hivebrain did mention "CRAM dots" very briefly but there's no explanation of what (or "where") they are (I think I originally thought they might be these "dots"). Not his fault though - not sure anyone ever asked until now :v:
Actually, no, you are exactly right, the link regarding the water dots is exactly the same the dots you see on the bottom of NTSC games, it's the exact same issue. The VDP displaying the boarder is the same as the VDP displaying the graphics during tile display, even though the VDP doesn't display graphics in the boarder, it is still reading from CRAM in order to draw that boarder colour while it's waiting for the TV to reach "real" V-blank, and writing to CRAM during H-blank is overwriting the storage colour. It is the exact same dots.